Fernanda Baião

PUC-RJ, Industrial Engineering Department, Faculty Member

Followers

Following

Co-authors

Public Views

Fernanda Baião is a Professor at the Dept. of Industrial Engineering at PUC-Rio, Brazil. Her research areas cover Conceptual Modeling and Ontologies, Data Science and BPM, with over 150 scientific publications. From 2004 to 2018 she was a Professor af the Federal University of the State of Rio de Janeiro, where she led a research group on Ontology Matching, and applying Cognitive Sciences and Natural Language Processing to Process Mining. From 2001 to 2004 she was a post-doc at COPPE/UFRJ, from where she obtained her D.Sc (2001) and her M.Sc (1997). During 2000 she visited the University of Wisconsin-Madison (USA). Her D.Sc Thesis proposed a pioneer solution for Designing Distributed Databases by integrating Optimization Meta-heuristics and Theory Revision techniques from the AI field.

Fernanda also developed a valuable expertise in transferring results from Academia to Industry. Particularly, she prospected and acted as a Technical Leader in R&D projects on BPM, Enterprise Architecture, Information Security and Data Mining in the areas of Oil & Gas Exploration and Production, Insurance Management and IT Service Management.
Supervisors: Marta Lima de Queirós Mattoso
Phone: +55-21-988807337

less

Interests

Uploads

Papers by Fernanda Baião

Aumentando a Expressividade Semântica na Modelagem de Regras de Negócio no contexto de Processos intensivos em Conhecimento(Increasing the Semantic Expressiveness of Business Rules Models in the context of Knowledge-intensive Processes)

ONTOBRAS, 2018

Business rules capture the essential restrictions that should be enforced in all possible worlds ... more Business rules capture the essential restrictions that should be enforced in all possible worlds of a domain, providing a more adequate paradigm for representing processes in which the need for flexibility is critical, known as knowledge-intensive processes. Traditional languages for representing process models typically do not handle flexible scenarios adequately, treating exceptions as a deviation in the process flow and ignoring other aspects of a knowledge-intensive process. This work proposes an approach for creating knowledge-intensive processes models with higher semantic expressiveness, focusing on the business rules perspective.

Download

A Knowledge-Based Perspective Of TheDistributed Design Of Object OrientedDatabases

WIT Transactions on Information and Communication Technologies, 1970

The performance of applications on Object Oriented Database Management Systems (OODBMSs) is stron... more The performance of applications on Object Oriented Database Management Systems (OODBMSs) is strongly affected by Distributed Design, which reduces irrelevant data accessed by applications and data exchange among sites. In an OO environment, the Distributed Design is a very complex task, and an open research problem. In this work we propose a knowledge based approach to the fragmentation phase of the distributed design of object oriented databases. In this approach, we will show a rule-based implementation of an analysis algorithm from our previous work and propose some ideas towards the use of Inductive Logic Programming (ILP) to perform a knowledge discovery/revision process using our set of rules as background knowledge. The objective of the work is to uncover some previously unknown issues to be considered in the distributed design process. Our main objective here is to show the viability of performing a revision process in order to obtain better and better fragmentation algorithms. We do not intend to propose the best fragmentation algorithm ever possible. We concentrate here on the process of revising a DDOODB algorithm through Knowledge Discovery techniques, rather than only obtaining a final optimal algorithm.

Download

OntoDW: An Approach for Extraction of Conceptualizations from Data Warehouses

ONTOBRAS, 2016

Business Intelligence (BI) fosters proper decision-making in organizations, mainly by providing t... more Business Intelligence (BI) fosters proper decision-making in organizations, mainly by providing the means to analyze historical data stored in repositories called Data Warehouses (DW). However, formal representation of which concepts are implemented in a DW rarely exists, which would be important to clarify and semantically describe the domain concepts behind the data stored in a DW, as well as the analytical concepts that are available for the BI tools. Examples of important pieces of knowledge that are frequently hidden into the DW are: which domain concepts are available as analysis perspectives (dimensions), how the domain concepts relate to each other, which metrics (facts) are available and what do they mean, which domain perspectives are considered for each metric and how metrics may be aggregated. On the other hand, one of the relevant uses of an ontology for the Computer Science area is as a codified artifact that formally represents a shared conceptualization about a universe of discourse. Therefore, ontologies can be used to represent both domain and analytical concepts codified and stored in a DW. However, extracting these concepts from an already-inproduction DW is not a trivial task, especially in medium and large organizations, often with tens of metrics and tens (even hundreds) of dimensions and potential aggregations. In this paper, we define a set of mapping rules from DW constructs to conceptual elements (concepts and relationships), towards automatically extracting an ontology codified in OWL. The proposal was successfully evaluated in a real scenario of a Brazilian financial institution.

Download

On the Identification and Representation of Ontology Correspondence Antipatterns

Semantic data integration and application interoperability are recognized challenges in distribut... more Semantic data integration and application interoperability are recognized challenges in distributed scenarios, and they have long being addressed by ontology alignment techniques, which try to find correspondences between entities that are semantically equivalent in distinct ontologies. However, ontology alignment is still a costly and difficult task, considering the existence of large-scale scenarios and complex domains. This work introduces the concept of a correspondence antipattern, which is essentially a set of generic correspondences that represent an incorrect alignment and can be used as a predefined template to help identify incorrect correspondences between two existing ontologies and thus refine ontology alignments. We also introduce a methodology that assists the identification and construction of ontology correspondence antipatterns, and evaluate it through a case study using datasets from the last three campaigns of OAEI, the Ontology Alignment Evaluation Initiative.

Download

A Measurement Ontology for Beliefs, Desires, Intentions and Feelings within Knowledge-intensive Processes

ONTOBRAS, 2018

In Knowledge-intensive Processes (KiPs) participants perform activities and collaborate with each... more In Knowledge-intensive Processes (KiPs) participants perform activities and collaborate with each other driven by their Beliefs, Desires, Intentions and Feelings (BDIFs). Measurement provides useful information for reaching conclusions and making decisions. However, it may present some particularities such as a difficulty to quantify intangible and subtle concepts like BDIFs. This paper presents an ontology that provides a way towards capturing and leveraging the intensity of BDIFs from their expression within conversations registered during KiP's execution. It was built based on the Measurement Ontology Pattern Language, the Speech Act Theory and the Knowledge-intensive Process Ontology. Its application is exemplified in a real life scenario.

Download

Alocação de Dados em Bancos de Dados Distribuídos

Brazilian Symposium on Databases, 2003

... Matheus Wildemberg, Melise MV Paula, Fernanda Baião, Marta Mattoso {mwild, mel, baiao, marta}... more

Analysis of COVID-19 under-reporting in Brazil/ Análise da subnotificação de COVID-19 no Brasil

Revista Brasileira De Terapia Intensiva, Jun 1, 2020

Foundational ontologies, ontology‐driven conceptual modeling, and their multiple benefits to data mining

WIREs Data Mining and Knowledge Discovery, 2021

For many years, the role played by domain knowledge in all stages of knowledge discovery has been... more For many years, the role played by domain knowledge in all stages of knowledge discovery has been recognized. However, the real‐world semantics embedded in data is often still not fully considered in traditional data mining methods. In this article, we argue that the quality of data mining results is directly related to the extent that they reflect important properties of real‐world entities represented therein. Analyzing and characterizing the nature of these entities is the very business of the area of formal ontology. We briefly elaborate on two particular types of artifacts produced by this area: foundational ontologies and ontology‐driven conceptual modeling languages grounded on them. We then elaborate on the benefits they can bring to several activities in a data mining process.This article is categorized under: Fundamental Concepts of Data and Knowledge > Knowledge Representation Fundamental Concepts of Data and Knowledge > Data Concepts

Outer-Tuning

Proceedings of the XV Brazilian Symposium on Information Systems, 2019

Database tuning is a crucial task to address the performance of information systems that deal wit... more Database tuning is a crucial task to address the performance of information systems that deal with a considerable amount of information stored in databases. Current tuning tools are very platform-specific and do not provide adequate support for the database administrator to reason about performance improvement suggestions. In this paper, we discuss several architectural and implementation decisions of Outer-Tuning, our framework that supports database tuning. Outer-Tuning follows a model-driven development and a modular architecture design, which enabled several benefits. This paper contributes with: (i) the architectural design model adopted in Outer-Tuning, which combines imperative and declarative programming; (ii) the discussions and steps to integrate several software components; and (iii) the actual framework implementation. We assess our framework with an experiment using the TPC-H benchmark. The results evidence that Outer-Tuning infers useful tuning actions and supports the DBA by providing a more semantic environment to create and adapt tuning heuristics using concepts closer to his/her domain, and also relevant information on the rationale of the tuning actions through a friendly web interface.

Towards a cluster-based approach for user participation in ontology maching

User participation is a promising approach for Ontology Matching; however, determining the most r... more

Download

Sociodemographic factors associated with COVID-19 in-hospital mortality in Brazil

Public Health, 2021

Objectives: The coronavirus disease 2019 (COVID-19) pandemic has highlighted inequalities in acce... more Objectives: The coronavirus disease 2019 (COVID-19) pandemic has highlighted inequalities in access to healthcare systems, increasing racial disparities and worsening health outcomes in these populations. This study analysed the association between sociodemographic characteristics and COVID-19 in-hospital mortality in Brazil. Study design: A retrospective analysis was conducted on quantitative reverse transcription polymerase chain reactioneconfirmed hospitalised adult patients with COVID-19 with a defined outcome (i.e. hospital discharge or death) in Brazil. Data were retrieved from the national surveillance system database (SIVEP-Gripe) between February 16 and August 8, 2020. Methods: Clinical characteristics, sociodemographic variables, use of hospital resources and outcomes of hospitalised adult patients with COVID-19, stratified by self-reported race, were investigated. The primary outcome was in-hospital mortality. The association between self-reported race and in-hospital mortality, after adjusting for clinical characteristics and comorbidities, was evaluated using a logistic regression model. Results: During the study period, Brazil had 3,018,397 confirmed COVID-19 cases and 100,648 deaths. The study population included 228,196 COVID-19epositive adult in-hospital patients with a defined outcome; the median age was 61 years, 57% were men, 35% (79,914) self-reported as Black/Brown and 35.4% (80,853) self-reported as White. The total in-hospital mortality was 37% (85,171/228,196). Black/ Brown patients showed higher in-hospital mortality than White patients (42% vs 37%, respectively), were admitted less frequently to the intensive care unit (ICU) (32% vs 36%, respectively) and used more invasive mechanical ventilation (21% vs 19%, respectively), especially outside the ICU (17% vs 11%, respectively). Black/Brown race was independently associated with high in-hospital mortality after adjusting for sex, age, level of education, region of residence and comorbidities (odds ratio ¼ 1.15; 95% confidence interval ¼ 1.09e1.22). Conclusions: Among hospitalised Brazilian adults with COVID-19, Black/Brown patients showed higher in-hospital mortality, less frequently used hospital resources and had potentially more severe conditions than White patients. Racial disparities in health outcomes and access to health care highlight the need to actively implement strategies to reduce inequities caused by the wider health determinants, ultimately leading to a sustainable change in the health system.

Download

Predicting drug sensitivity of cancer cells based on DNA methylation levels

PLOS ONE, 2021

Cancer cell lines, which are cell cultures derived from tumor samples, represent one of the least... more Cancer cell lines, which are cell cultures derived from tumor samples, represent one of the least expensive and most studied preclinical models for drug development. Accurately predicting drug responses for a given cell line based on molecular features may help to optimize drug-development pipelines and explain mechanisms behind treatment responses. In this study, we focus on DNA methylation profiles as one type of molecular feature that is known to drive tumorigenesis and modulate treatment responses. Using genome-wide, DNA methylation profiles from 987 cell lines in the Genomics of Drug Sensitivity in Cancer database, we used machine-learning algorithms to evaluate the potential to predict cytotoxic responses for eight anti-cancer drugs. We compared the performance of five classification algorithms and four regression algorithms representing diverse methodologies, including tree-, probability-, kernel-, ensemble-, and distance-based approaches. We artificially subsampled the data ...

Download

Characterisation of the first 250 000 hospital admissions for COVID-19 in Brazil: a retrospective analysis of nationwide data

The Lancet Respiratory Medicine, 2021

Background Most low-income and middle-income countries (LMICs) have little or no data integrated ... more Background Most low-income and middle-income countries (LMICs) have little or no data integrated into a national surveillance system to identify characteristics or outcomes of COVID-19 hospital admissions and the impact of the COVID-19 pandemic on their national health systems. We aimed to analyse characteristics of patients admitted to hospital with COVID-19 in Brazil, and to examine the impact of COVID-19 on health-care resources and in-hospital mortality. We did a retrospective analysis of all patients aged 20 years or older with quantitative RT-PCR (RT-qPCR)confirmed COVID-19 who were admitted to hospital and registered in SIVEP-Gripe, a nationwide surveillance database in Brazil, between Feb 16 and Aug 15, 2020 (epidemiological weeks 8-33). We also examined the progression of the COVID-19 pandemic across three 4-week periods within this timeframe (epidemiological weeks 8-12, 19-22, and 27-30). The primary outcome was in-hospital mortality. We compared the regional burden of hospital admissions stratified by age, intensive care unit (ICU) admission, and respiratory support. We analysed data from the whole country and its five regions: North, Northeast, Central-West, Southeast, and South. Between Feb 16 and Aug 15, 2020, 254 288 patients with RT-qPCR-confirmed COVID-19 were admitted to hospital and registered in SIVEP-Gripe. The mean age of patients was 60 (SD 17) years, 119 657 (47%) of 254 288 were aged younger than 60 years, 143 521 (56%) of 254 243 were male, and 14 979 (16%) of 90 829 had no comorbidities. Case numbers increased across the three 4-week periods studied: by epidemiological weeks 19-22, cases were concentrated in the North, Northeast, and Southeast; by weeks 27-30, cases had spread to the Central-West and South regions. 232 036 (91%) of 254 288 patients had a defined hospital outcome when the data were exported; in-hospital mortality was 38% (87 515 of 232 036 patients) overall, 59% (47 002 of 79 687) among patients admitted to the ICU, and 80% (36 046 of 45 205) among those who were mechanically ventilated. The overall burden of ICU admissions per ICU beds was more pronounced in the North, Southeast, and Northeast, than in the Central-West and South. In the Northeast, 1545 (16%) of 9960 patients received invasive mechanical ventilation outside the ICU compared with 431 (8%) of 5388 in the South. In-hospital mortality among patients younger than 60 years was 31% (4204 of 13 468) in the Northeast versus 15% (1694 of 11 196) in the South. Interpretation We observed a widespread distribution of COVID-19 across all regions in Brazil, resulting in a high overall disease burden. In-hospital mortality was high, even in patients younger than 60 years, and worsened by existing regional disparities within the health system. The COVID-19 pandemic highlights the need to improve access to high-quality care for critically ill patients admitted to hospital with COVID-19, particularly in LMICs.

Download

Progression of confirmed COVID-19 cases after the implementation of control measures

Revista Brasileira de Terapia Intensiva, 2020

We study an optomechanical cavity, in which a buckled suspended beam serves as a mirror. The mech... more We study an optomechanical cavity, in which a buckled suspended beam serves as a mirror. The mechanical resonance frequency of the beam obtains a minimum value near the buckling temperature. Contrary to the common case, in which self-excited oscillations of the suspended mirror are optically induced by injecting blue detuned laser light, in our case self-excited oscillations are observed with red detuned light. These observations are attributed to a retarded thermal (i.e. bolometric) force acting on the buckled mirror in the inwards direction (i.e. towards to other mirror). With relatively high laser power other interesting effects are observed including period doubling of self-excited oscillations and intermode coupling.

Download

Alignment Patterns Based on Unified Foundational Ontology

Ontology alignment is the process of finding related entities in different ontologies. In this co... more Ontology alignment is the process of finding related entities in different ontologies. In this context, precise and explicit representation of conceptualizations is essential for reaching semantic integration to ensure that only data related to the same (or sufficiently similar) real-world entity are merged. Foundational ontologies describe general concepts independent of a domain and precisely define meta-properties so as to make the semantics of each concept in the ontology explicit. In this paper we show how the use of OntoUML, a conceptual modeling language based on Unified Foundational Ontology, allows the application of alignment patterns and exemplify how this approach may improve precision, recall and refine the type of the alignment.

Download

Towards Planning Scientific Experiments through Declarative Model Discovery in Provenance Data

2014 IEEE 10th International Conference on e-Science, 2014

Data provenance is the process of managing a collection of metadata that catalogs the origin and ... more Data provenance is the process of managing a collection of metadata that catalogs the origin and history of data. In scientific workflows, this metadata assists scientists and domain specialists in several tasks, including the reproduction of scientific experiments and planning of new scenarios to be experimented. However, the amount of provenance data generated from scientific workflow executions can grow through time, becoming infeasible for scientists to manually evaluate them. Thus, mechanisms for automatically extracting knowledge from provenance data and presenting them to the user are demanding. Due to the diversity and flexibility inherent to scientific experimentation scenarios, declarative models are potentially adequate. In this work, we propose to apply techniques for learning a declarative model from provenance data generated by scientific workflows, from which the domain specialist will be able to plan future scenarios for his/her scientific experiment. The proposed solution is illustrated in a case study of a scientific experiment on ontology matching, which is a data-intensive strategy that is required to solve the problem of information integration in several areas of knowledge.

A Method for Discovering the Relevance of External Context Variables to Business Processes

Proceedings of the International Conference on Knowledge Management and Information Sharing, 2011

Organizations have been demanded to efficiently detect and respond to changes in their environmen... more Organizations have been demanded to efficiently detect and respond to changes in their environment, which depends on its ability to adapt their business processes. Taking internal and external environment variables into account enables to address issues, such as, how a business process was executed last time the country experienced a similar economic scenario; whether that process execution brought positive results or not; which were the external environmental reasons that provoked changes in previous process executions. These environmental variables are typically referred in the literature as the context of the process. In this paper, we propose a method to identify and prioritize external variables that impact the execution of specific activities of a process. The proposed method applies competitive intelligence concepts and data mining techniques, and was evaluated in a case study.

Download

Towards a context-based representation of the dynamicity perspective in knowledge-intensive processes

A Knowledge-intensive process (KIP) is characterized as a bag of activities based on knowledge in... more A Knowledge-intensive process (KIP) is characterized as a bag of activities based on knowledge intensive acquisition and manipulation. A KIP differentiates from traditional business processes mainly because of the dynamism or the variability of event flows among the process instances. Thus, traditional business process modeling notations do not adequately represent the broad spectrum of flows that occur in each KIP instance, making them even more challenging to be understood or managed. Some approaches to represent a KIP were defined in the literature, but representing the KIP variability in a cognitively efficient way is still an open issue. This paper presents the graphical notation KIPN-C, which proposes diagrams to represent variability of a KIP relating it to the context of its instances.

Applying Multi-Level Typing to Model Knowledge-Intensive Processes

ONTOBRAS, 2017

Modeling KnowledgeIntensive Processes (KIP) is very important for understanding critical scenario... more Modeling KnowledgeIntensive Processes (KIP) is very important for understanding critical scenarios in current organizations. KIPO (KnowledgeIntensive Process Ontology) is an ontology wellfounded, semantically rich conceptualization of KIP. However, it is difficult to distinguish instances and models in KIP. Our goal is to propose an application with the notion of multilevel conceptual modeling for representing elements with multiple classification level.

Download

XCuteKIP: Support for Knowledge Intensive Process Activities

Lecture Notes in Computer Science, 2015

The century of information puts the organizational processes of which the work highlights knowled... more The century of information puts the organizational processes of which the work highlights knowledge as the main asset into perspective, as well as an engine generator of competitive advantage. The nature of processes that include intensive use of knowledge points out several features that can be exploited to assist in adding value to products and services supported by organizations. These processes are known as Knowledge-intensive Processes. This article presents the XcuteKIP architecture that aims to support the participants of these processes. Particularly, the proposal assists in the execution of knowledge-intensive collaborative activities by providing semi-automatic recommendations of collaborative services. The approach applies the concepts of a service-oriented architecture subsidized by WGWSOA and a semantic modeling Ontology supported by KIPO.

Fernanda Baião

Uploads

Papers by Fernanda Baião

Log In